Goodman, Leo A., 2007, "Otis Dudley Duncan, Quantitative Sociologist Par Excellence: Path Analysis, Loglinear Methods, and Rasch Model," Research in Social Stratification and Mobility, 25: 129-39.
From the very beginning of his work on path analysis, Dudley was acutely aware of its limitations as a method for discovering causes. In his classic article on path analysis, published in 1966, he stated that “path analysis focuses on the problem of interpretation and does not purport to be a method for discovering causes;” and in the 1967 Blau-Duncan book, there is the following statement: “Sociologists have only recently begun to appreciate how stringent are the logical requirements that must be met if discussion of causal processes is to go beyond mere impressionism and vague verbal formulations. We are a long way from being able to make causal inferences with confdence, and schemes of the kind presented here had best be regarded as crude first approximations to adequate causal models.”
From the early 1970s on, Dudley no longer applied path analysis or structural equation models in any of the many diverse research projects that he carried out over those years and, in his 1975 text on structural equation models, he omitted all empirical examples.